In response to the Request for Applications for Applications #DK 01-009, entitled "Mouse Models of Diabetic Complications Consortium," the Mouse Diabetic Nephropathy and Neuropathy United (MDNNU) at the University of Michigan will develop models of diabetic neuropathy and neuropathy in mice via state-of-the-art mouse engineering. The technology that emphasizes cell-type specific knockout strategies manipulating genes that will predispose diabetic mice to these two diabetic complications. The unit will comprise a number of investigators at the University of Michigan and two investigators at Johns Hopkins University, who have wide-ranging experience and expertise in mouse genetic engineering, mouse model investigation, and diabetic phenotyping. Rodent models of diabetes fail to develop changes that closely resemble human diabetic nephropathy or neuropathy. The reasons for the resistance to full-blown complications are likely manifold, but may include absence of important genetic susceptibility genes and, more simply, the life-span of rodents that is too limited to allow full development of complications. Our general strategic approach to this dilemma is to accelerate the injury of diabetes is predisposing critical cells in both the renal glomerulus and peripheral nerve to injury. Our strategy will be simpler: to alter expression of gene products known to be associated with cellular injury and which appear to be involved in diabetic complications, in order to directly prime glomerular and peripheral nerve cells in a cell type- and temporally-specific manner for injury once diabetes occurs. If successful, these models will show rapid progression of diabetic injury, will be easy and inexpensive to study because of this rapid and reproducible downhill course, and will provide excellent models for evaluation of potential therapeutic interventions. To best address the goals of this initiative to develop novel models of diabetic nephropathy and neuropathy, we will pursue the following aims: 1. Induce diabetes in mice which are null, heterozygote, or wild type for mitochondrial superoxide dismutase (SOD2). 2. Induce diabetes in mice which are null, heterozygote, or wild type for gamma glutamate cysteine ligase (GCLC). 3. Phenotype SOD2 and GCLC diabetic and control mice for nephopathy and neuropathy. 4. Create tetracycline-regulatable Cre-lox conditional knockout mice specific for peripheral neuron and/or podocyte-specific knockout of SOD2 and GCLC. 5. Fully develop and phenotype the cell-type specific conditional knockouts as suggested by phenotypic results in conventional knockouts.